A digital twin method, device, equipment and storage medium for market data

By classifying massive market data and building entity relationship diagrams and generating voucher data sets with the location of market entities, the problems of inefficient and large errors in digital twins in the existing technology are solved, and accurate assessment of market risks and markets are achieved Stability guarantee.

CN118657388BActive Publication Date: 2024-11-05GUIZHOU UNIVERSITY OF FINANCE AND ECONOMICS +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411147799.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-21
Publication Date
2024-11-05
Estimated Expiration
2044-08-21

AI Technical Summary

Technical Problem

The existing technology is inefficient and has large errors when digitally twinning massive market data, ignoring additional information of market data, making it difficult to accurately reflect market risks.

Method used

By obtaining the historical market data of the target area, classifying it in combination with the classification model, building an entity relationship diagram of the market entity data set and market behavior data, using the market entity location to build a coordinate system, compute the mapping values ​​of nodes, edges and paths, and generate credential data and evaluate market risks based on credential data sets.

Benefits of technology

It improves the digital twin accuracy of market data, ensures the accuracy of market risk assessment, enhances the correlation with actual entities, and ensures market stability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118657388B_ABST
    Figure CN118657388B_ABST
Patent Text Reader

Abstract

The present invention discloses a digital twin method and device, equipment and storage medium for market data, and relates to the field of data processing technology. The present invention obtains historical market data of a target area, and obtains a market subject data set and a market subject origin through a classification model, and efficiently obtains a large number of different types of market subjects; then, by constructing a first entity relationship diagram of the market subject data set and the market behavior data, a preliminary connection between the market subject and the market behavior is established, and the coordinates of the historical market data are constructed in combination with the market subject position of the target area, thereby constructing a voucher data set for the target area; the voucher data set constructed with actual location data enhances the correlation with the actual entity, thereby making the digital twin of the historical market data more accurate, ensuring that the actual market risk assessment of the target area based on the voucher data set is more accurate, and ensuring the market stability of the region.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a digital twin method and device, equipment and storage medium for market data. Background Art

[0002] Nowadays, with the rapid development of social informatization, digital twinning of data has become an important development direction. Digital twinning of data refers to the virtual mapping of key features and operating data of physical entities in the real world. As a special type of data, market data is generated faster and faster with the continuous advancement of technology and the continuous expansion of the market, and the amount is also increasing.

[0003] However, since market data not only involves data within a certain region, but also involves data interaction in multiple regions, and includes massive amounts of physical data, it is more difficult to create digital twins of market data. In existing research on digital twins of market data, only mathematical tools are usually used for digital twins, but this method ignores a lot of additional information about market data, and when faced with massive amounts of data, using only mathematical tools for digital twins has problems such as low efficiency and large errors. Therefore, a new method is urgently needed to solve the defects of existing technologies. Summary of the invention

[0004] The present invention provides a digital twin method and apparatus, equipment and storage medium for market data to solve the technical problem of how to accurately perform digital twinning on massive market data.

[0005] In order to solve the above technical problems, an embodiment of the present invention provides a digital twin method of market data, including:

[0006] Obtain historical market data of the target area and classify it in combination with the preset classification model to obtain the market entity data set and market entity origin;

[0007] According to a preset search algorithm, combined with the market subject data set, the market behavior data corresponding to each market subject is obtained from the historical market data, and a first entity relationship diagram between the market subject data set and the market behavior data is constructed;

[0008] Obtain the location data of each market entity in the target area; take the location of the market entity origin as the coordinate origin, and combine the location data of each market entity to construct the market entity location coordinate system of the target area to obtain the coordinate information of each market entity; calculate the mapping value of each node, edge and path in the first entity relationship diagram according to the coordinate information of each market entity; obtain the timestamp information and behavior type of each market behavior, and combine the mapping value of each node and path in the first entity relationship diagram to construct the voucher data set of the target area;

[0009] According to the voucher data set, the market risk of the target area is evaluated; wherein, the market risk of the target area is evaluated according to the voucher data set, specifically comprising: classifying the voucher data set according to the voucher data set in combination with a preset classification method; dividing the voucher data set into blocks and performing statistics according to the classification results to obtain a number of risk scores corresponding to each first voucher value in the voucher data set; obtaining a preset risk assessment standard, and evaluating the market risk of the target area in combination with the risk score.

[0010] It can be understood that, compared with the prior art, the present invention obtains the historical market data of the target area, and obtains the market entity data set and the market entity origin through the classification model, making full use of the classification model's efficient processing capabilities for massive data, thereby efficiently obtaining a large number of different types of market entities; then, by constructing a first entity relationship diagram between the market entity data set and the market behavior data, a preliminary connection between the market entity and the market behavior is established, and combined with the market entity location of the target area, the coordinates of the historical market data are constructed, thereby constructing a voucher data set for the target area; the voucher data set constructed with actual location data enhances the correlation with the actual entity, thereby making the digital twin of the historical market data more accurate, ensuring that the actual market risk assessment of the target area based on the voucher data set is more accurate, and ensuring the market stability of the region.

[0011] As a preferred solution, the historical market data of the target area is obtained, and classified in combination with a preset classification model to obtain a market entity data set and a market entity origin, which specifically includes:

[0012] Obtain historical market data of the target area, clean the historical market data and extract market entity types to obtain initial historical market data and market entity types;

[0013] Inputting the initial historical market data into a preset classification model, and combining the market subject types to obtain a plurality of classification sets output by the classification model;

[0014] According to the market entity type, market entities are extracted from each classification set to obtain a market entity data set;

[0015] Calculate the feature points of each classification set separately to obtain several feature points;

[0016] According to the preset origin calculation method, the market entity origin is calculated in combination with the several characteristic points.

[0017] This preferred solution can filter out data that does not conform to the actual market by preprocessing historical market data; relying on the excellent processing ability of the classification model for massive data, it can efficiently classify and divide massive market data.

[0018] As a preferred solution, the method of obtaining the market behavior data corresponding to each market entity from the historical market data according to the preset search algorithm and combining the market entity data set, and constructing a first entity relationship diagram between the market entity data set and the market behavior data, specifically includes:

[0019] According to the preset search algorithm, combined with each market entity in the market entity data set, the behavior data corresponding to each market entity is obtained from the historical market data, and a behavior information table corresponding to each market entity is constructed;

[0020] According to the knowledge graph algorithm, a node is constructed for each market entity in turn, and an initial entity relationship diagram of the market entity data set and the market behavior data is constructed in combination with the behavior information table of each market entity;

[0021] According to the depth-first algorithm, the initial entity relationship graph is traversed to find several paths;

[0022] Each path is detected in turn. When the starting node and the end node of the path are the same, the path is removed from the initial entity relationship graph until there is no path with the same starting node and end node in the initial entity relationship graph, and the optimization of the initial entity relationship graph is completed to obtain the first entity relationship graph.

[0023] This preferred solution obtains market behavior data through a search algorithm, and relies on a knowledge graph algorithm to construct an entity relationship diagram between market entities and market behavior data. The entity relationship diagram can intuitively and clearly reflect the behavioral relationship between market entities; by traversing the entity relationship diagram and eliminating paths, data that does not conform to actual market rules can be removed, making the entity relationship diagram more in line with actual market rules.

[0024] As a preferred solution, the calculation of the mapping value of each node, edge and path in the first entity relationship diagram according to the coordinate information of each market entity specifically includes:

[0025] Add the horizontal coordinate value and the vertical coordinate value of each market entity and take the absolute value to obtain the mapping value of the node corresponding to the market entity;

[0026] Get the coordinate information of the nodes at both ends of each edge, calculate the Euclidean distance between the nodes at both ends, and get the mapping value of the edge;

[0027] According to a preset path mapping function, the mapping value of each node and edge is combined to obtain the mapping value of each path in the first entity relationship graph;

[0028] The preset path mapping function includes:

[0029] ,

[0030] in, is the mapping value of the path, is the number of edges in the path, is the mapping value of each edge, is the sum of the mapping values ​​of the nodes and edges on the path, is the mapping value of the starting node of the path, is the mapping value of the end node of the path, and It refers to performing a base-10 logarithm operation on the mapping values ​​of the starting node and the end node of the path.

[0031] This preferred solution obtains the mapping value of each path through mapping value assignment. Through the mapping value, the behavior of market entities can be matched with actual market demand, thereby improving the accuracy of the digital twin of market data.

[0032] As a preferred solution, the acquisition of the timestamp information and behavior type of each market behavior, combined with the mapping value of each node and path in the first entity relationship diagram, to construct a voucher data set for the target area specifically includes:

[0033] Obtaining timestamp information of each market behavior, converting the format of the timestamp to obtain the first timestamp of each market behavior;

[0034] According to a preset label assignment method, assign a value to the behavior type of each market behavior, so that each market behavior has a behavior type value;

[0035] According to the behavior type value of each market behavior, the first behavior type value of each path is obtained;

[0036] Generate an initial voucher value corresponding to each path according to a preset voucher data generation function, combining the first timestamp of each market behavior and the mapping value of each node and path in the first entity relationship graph;

[0037] The preset credential data generation function includes:

[0038] ,

[0039] in, is the credential data corresponding to the path. is the mapping value of the starting node of the path, is the mapping value of the end node of the path, is a natural constant, is the mapping value of the path, The first timestamp of the market behavior corresponding to the path; To perform a second root operation on a value;

[0040] According to a preset splicing method, the initial credential value of the path and the first behavior type value are spliced ​​to obtain the first credential value of the path, and then the credential data set of the target area is obtained.

[0041] This preferred solution generates a credential value through the behavior type of market behavior, the mapping value of nodes and paths, and the timestamp, so that the credential value can fully integrate the basic demand information of actual market behavior and improve the correlation between the virtual credential value and the actual physical market behavior.

[0042] As a preferred solution, the market risk of the target area is evaluated based on the voucher data set, specifically including:

[0043] According to the voucher data set, combined with a preset classification method, the voucher data set is classified;

[0044] According to the classification results, the voucher data set is divided into blocks and counted to obtain a number of risk scores corresponding to each first voucher value in the voucher data set;

[0045] Obtain preset risk assessment criteria, and evaluate the market risk of the target area in combination with the risk score.

[0046] This preferred solution obtains several risk scores by classifying and dividing the voucher data set into blocks for statistics, and evaluates the market risk of the region based on the risk assessment standards, so that the risk assessment is more in line with the actual market operation standards, thereby improving the assessment accuracy of the market analysis of the target area.

[0047] Accordingly, an embodiment of the present invention provides a digital twin device for market data, including: a market subject acquisition module, an entity relationship diagram construction module, a voucher data calculation module, and a risk assessment module;

[0048] The market subject acquisition module is used to acquire the historical market data of the target area, and classify it in combination with a preset classification model to obtain a market subject data set and a market subject origin;

[0049] The entity relationship diagram construction module is used to obtain the market behavior data corresponding to each market entity from the historical market data according to the preset search algorithm and in combination with the market entity data set, and to construct a first entity relationship diagram of the market entity data set and the market behavior data;

[0050] The voucher data calculation module is used to obtain the location data of each market entity in the target area; using the location of the market entity origin as the coordinate origin, combined with the location data of each market entity, construct the market entity location coordinate system of the target area, and obtain the coordinate information of each market entity; according to the coordinate information of each market entity, calculate the mapping value of each node, edge and path in the first entity relationship diagram; obtain the timestamp information and behavior type of each market behavior, and construct the voucher data set of the target area in combination with the mapping value of each node and path in the first entity relationship diagram;

[0051] The risk assessment module is used to assess the market risk of the target area based on the voucher data set; wherein, the assessing the market risk of the target area based on the voucher data set specifically includes: classifying the voucher data set based on the voucher data set in combination with a preset classification method; dividing the voucher data set into blocks and performing statistics based on the classification results to obtain a number of risk scores corresponding to each first voucher value in the voucher data set; obtaining a preset risk assessment standard, and assessing the market risk of the target area in combination with the risk score.

[0052] It can be understood that, compared with the prior art, the present device obtains the historical market data of the target area, and obtains the market entity data set and the market entity origin through the classification model, making full use of the classification model's efficient processing capabilities for massive data, thereby efficiently obtaining a large number of different types of market entities; then, by constructing a first entity relationship diagram between the market entity data set and the market behavior data, a preliminary connection between the market entity and the market behavior is established, and combined with the market entity location in the target area, the coordinates of the historical market data are constructed, thereby constructing a voucher data set for the target area; the voucher data set constructed with actual location data enhances the correlation with the actual entity, thereby making the digital twin of the historical market data more accurate, ensuring that the actual market risk assessment of the target area based on the voucher data set is more accurate, and ensuring the market stability of the region.

[0053] Accordingly, an embodiment of the present invention provides a terminal device, including:

[0054] one or more processors;

[0055] A memory, coupled to the processor, for storing one or more programs;

[0056] When the one or more programs are executed by the one or more processors, the one or more processors implement the digital twin method of market data as described above.

[0057] Accordingly, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the digital twin method of market data as described above. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 : A flowchart of a method for digital twinning of market data provided by an embodiment of the present invention;

[0059] Figure 2 : A schematic diagram of the structure of a digital twin device for market data provided by an embodiment of the present invention;

[0060] Among them, the figure numbers of the drawings in the specification are as follows: 201: market entity acquisition module; 202: entity relationship diagram construction module; 203: voucher data calculation module; 204: risk assessment module. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] Embodiment 1

[0063] Please refer to Figure 1 , which is a step flow chart of a digital twin method for market data provided by an embodiment of the present invention, including steps S101-S104.

[0064] Step S101: Acquire historical market data of the target area, and classify it in combination with a preset classification model to obtain a market entity data set and a market entity origin.

[0065] In this embodiment, the acquisition of historical market data of the target area and classification based on a preset classification model to obtain a market entity data set and a market entity origin specifically includes:

[0066] Obtain historical market data of the target area, clean the historical market data and extract market entity types to obtain initial historical market data and market entity types;

[0067] Inputting the initial historical market data into a preset classification model, and combining the market subject types to obtain a plurality of classification sets output by the classification model;

[0068] According to the market entity type, market entities are extracted from each classification set to obtain a market entity data set;

[0069] Calculate the feature points of each classification set separately to obtain several feature points;

[0070] According to the preset origin calculation method, the market entity origin is calculated in combination with the several characteristic points.

[0071] By preprocessing historical market data, data that does not conform to the actual market can be filtered out; relying on the classification model's excellent processing capabilities for massive data, massive market data can be classified and divided efficiently.

[0072] In an optional embodiment, the preset classification model refers to a support vector machine algorithm (SVM), which uses the SVM algorithm to combine the market entity types to classify the market entity data in the historical market data according to the market entity types to obtain multiple classification sets; each classification set includes several market entities of the same type.

[0073] In an optional embodiment, the characteristic points of each classification set are calculated respectively to obtain a plurality of characteristic points; and the market entity origin is calculated according to a preset origin calculation method in combination with the plurality of characteristic points, specifically including:

[0074] The support vector of each classification set is used as the feature point of each classification set;

[0075] According to the actual geographical location of each feature point, a two-dimensional coordinate system is constructed to calculate the position of the center point and the distance between each feature point and the center point;

[0076] Select the feature point with the smallest distance from the center point as the origin of the market entity.

[0077] Step S102: According to a preset search algorithm and in combination with the market subject data set, the market behavior data corresponding to each market subject is obtained from the historical market data, and a first entity relationship diagram of the market subject data set and the market behavior data is constructed.

[0078] In this embodiment, the market behavior data corresponding to each market entity is obtained from the historical market data according to the preset search algorithm and in combination with the market entity data set, and a first entity relationship diagram between the market entity data set and the market behavior data is constructed, specifically including:

[0079] According to the preset search algorithm, combined with each market entity in the market entity data set, the behavior data corresponding to each market entity is obtained from the historical market data, and a behavior information table corresponding to each market entity is constructed;

[0080] According to the knowledge graph algorithm, a node is constructed for each market entity in turn, and an initial entity relationship diagram of the market entity data set and the market behavior data is constructed in combination with the behavior information table of each market entity;

[0081] According to the depth-first algorithm, the initial entity relationship graph is traversed to find several paths;

[0082] Each path is detected in turn. When the starting node and the end node of the path are the same, the path is removed from the initial entity relationship graph until there is no path with the same starting node and end node in the initial entity relationship graph, and the optimization of the initial entity relationship graph is completed to obtain the first entity relationship graph.

[0083] Market behavior data is obtained through search algorithms, and an entity relationship diagram between market entities and market behavior data is constructed by relying on knowledge graph algorithms. The entity relationship diagram can intuitively and clearly reflect the behavioral relationship between market entities; by traversing the entity relationship diagram and eliminating paths, data that does not conform to actual market rules can be removed, making the entity relationship diagram more in line with actual market rules.

[0084] Step S103: construct the coordinates of the historical market data according to the market entity location data and the market entity origin of the target area, and construct the voucher data set of the target area in combination with the first entity relationship diagram and the market behavior data.

[0085] In this embodiment, the coordinates of the historical market data are constructed based on the market subject location data and the market subject origin of the target area, and the voucher data set of the target area is constructed in combination with the first entity relationship diagram and the market behavior data, specifically including:

[0086] Obtain location data for each market player in the target area;

[0087] Taking the origin of the market entity as the coordinate origin, combined with the location data of each market entity, the market entity location coordinate system of the target area is constructed to obtain the coordinate information of each market entity;

[0088] Calculate the mapping value of each node, edge and path in the first entity relationship graph according to the coordinate information of each market entity;

[0089] The timestamp information and behavior type of each market behavior are obtained, and the voucher data set of the target area is constructed by combining the mapping values ​​of each node and path in the first entity relationship diagram.

[0090] By constructing the mapping value of the entity relationship diagram based on the location information of the market entities in the target area, and combining the timestamp and behavior type of the market behavior to construct a credential dataset, the credential dataset can accurately express the relationship between physical entities and virtual data entities in the real world.

[0091] In this embodiment, the mapping values ​​of each node, edge, and path in the first entity relationship diagram are calculated according to the coordinate information of each market entity, and the voucher data set of the target area is constructed in combination with the timestamp information of each market behavior, which specifically includes:

[0092] Add the horizontal coordinate value and the vertical coordinate value of each market entity and take the absolute value to obtain the mapping value of the node corresponding to the market entity;

[0093] Get the coordinate information of the nodes at both ends of each edge, calculate the Euclidean distance between the nodes at both ends, and get the mapping value of the edge;

[0094] According to a preset path mapping function, the mapping value of each node and edge is combined to obtain the mapping value of each path in the first entity relationship graph;

[0095] The preset path mapping function includes:

[0096] ,

[0097] in, is the mapping value of the path, is the number of edges in the path, is the mapping value of each edge, is the sum of the mapping values ​​of the nodes and edges on the path, is the mapping value of the starting node of the path, is the mapping value of the end node of the path, and It refers to performing a base-10 logarithm operation on the mapping values ​​of the starting node and the end node of the path.

[0098] By assigning mapping values, we get the mapping value of each path. Through the mapping value, the behavior of market entities can be matched with actual market demand, thereby improving the accuracy of the digital twin of market data.

[0099] In this embodiment, the acquisition of the timestamp information and behavior type of each market behavior, combined with the mapping value of each node and path in the first entity relationship diagram, to construct a voucher data set for the target area specifically includes:

[0100] Obtaining timestamp information of each market behavior, converting the format of the timestamp to obtain the first timestamp of each market behavior;

[0101] According to a preset label assignment method, assign a value to the behavior type of each market behavior, so that each market behavior has a behavior type value;

[0102] According to the behavior type value of each market behavior, the first behavior type value of each path is obtained;

[0103] Generate an initial voucher value corresponding to each path according to a preset voucher data generation function, combining the first timestamp of each market behavior and the mapping value of each node and path in the first entity relationship graph;

[0104] The preset credential data generation function includes:

[0105] ,

[0106] in, is the credential data corresponding to the path. is the mapping value of the starting node of the path, is the mapping value of the end node of the path, is a natural constant, is the mapping value of the path, The first timestamp of the market behavior corresponding to the path; To perform a second root operation on a value;

[0107] According to a preset splicing method, the initial credential value of the path and the first behavior type value are spliced ​​to obtain the first credential value of the path, and then the credential data set of the target area is obtained.

[0108] The credential value is generated through the behavior type of market behavior, the mapping value of nodes and paths, and the timestamp, so that the credential value can fully integrate the basic demand information of actual market behavior and improve the correlation between the virtual credential value and the actual physical market behavior.

[0109] Step S104: Evaluate the market risk of the target area based on the voucher data set.

[0110] In this embodiment, the market risk of the target area is evaluated based on the voucher data set, specifically including:

[0111] According to the voucher data set, combined with a preset classification method, the voucher data set is classified;

[0112] According to the classification results, the voucher data set is divided into blocks and counted to obtain a number of risk scores corresponding to each first voucher value in the voucher data set;

[0113] Obtain preset risk assessment criteria, and evaluate the market risk of the target area in combination with the risk score.

[0114] By classifying and dividing the voucher data set into blocks for statistics, we obtained several risk scores, and evaluated the market risk of the region based on the risk assessment standards, making the risk assessment more in line with the actual market operation standards, thereby improving the assessment accuracy of the market analysis of the target area.

[0115] In an optional embodiment, the evaluation of the credential data set based on a preset classification method refers to classification based on the behavior type value in each credential value in the credential data set; the block statistics are based on the classification results, a corresponding benchmark value is assigned to each credential value, and the benchmark value is used as the risk score of the credential data.

[0116] In an optional embodiment, obtaining a preset risk assessment standard and assessing the market risk of the target area in combination with the risk score may include:

[0117] Obtain a preset security value range, and when the risk score of the credential data is not within the security value range, mark the credential data as a risky credential;

[0118] The proportion of statistical risk vouchers to the voucher data set, when the proportion exceeds 10%, it means that the market risk of the target area is extremely high; when the proportion is between 5% and 10%, it means that the market risk of the target area is relatively small; when the proportion is less than 5%, it means that the market risk of the target area is extremely small.

[0119] This embodiment obtains historical market data of the target area, and obtains the market entity data set and the market entity origin through the classification model, making full use of the classification model's efficient processing capabilities for massive data, thereby efficiently obtaining a large number of different types of market entities; then, by constructing a first entity relationship diagram between the market entity data set and the market behavior data, a preliminary connection between the market entity and the market behavior is established, and combined with the market entity location of the target area, the coordinates of the historical market data are constructed, thereby constructing a voucher data set for the target area; the voucher data set constructed with actual location data enhances the correlation with the actual entity, thereby making the digital twin of the historical market data more accurate, ensuring that the actual market risk assessment of the target area based on the voucher data set is more accurate, and ensuring the market stability of the region.

[0120] Embodiment 2

[0121] Please refer to Figure 2 , which is a structural schematic diagram of a digital twin device of market data provided by an embodiment of the present invention, including: a market subject acquisition module 201, an entity relationship diagram construction module 202, a voucher data calculation module 203 and a risk assessment module 204;

[0122] The market entity acquisition module 201 is used to acquire historical market data of the target area, and classify it in combination with a preset classification model to obtain a market entity data set and a market entity origin.

[0123] In this embodiment, the market subject acquisition module 201 includes: a market subject acquisition unit,

[0124] The market subject acquisition unit is used to acquire historical market data of the target area, and perform data cleaning and market subject type extraction on the historical market data to obtain initial historical market data and market subject types;

[0125] Inputting the initial historical market data into a preset classification model, and combining the market subject types to obtain a plurality of classification sets output by the classification model;

[0126] According to the market entity type, market entities are extracted from each classification set to obtain a market entity data set;

[0127] Calculate the feature points of each classification set separately to obtain several feature points;

[0128] According to the preset origin calculation method, the market entity origin is calculated in combination with the several characteristic points.

[0129] The entity relationship diagram construction module 202 is used to obtain the market behavior data corresponding to each market entity from the historical market data according to a preset search algorithm and in combination with the market entity data set, and to construct a first entity relationship diagram of the market entity data set and the market behavior data.

[0130] In this embodiment, the entity relationship diagram construction module 202 includes: an entity relationship diagram construction unit;

[0131] The entity relationship diagram construction unit is used to obtain the behavior data corresponding to each market entity from the historical market data according to the preset search algorithm and in combination with each market entity in the market entity data set, and to construct a behavior information table corresponding to each market entity;

[0132] According to the knowledge graph algorithm, a node is constructed for each market entity in turn, and an initial entity relationship diagram of the market entity data set and the market behavior data is constructed in combination with the behavior information table of each market entity;

[0133] According to the depth-first algorithm, the initial entity relationship graph is traversed to find several paths;

[0134] Each path is detected in turn. When the starting node and the end node of the path are the same, the path is removed from the initial entity relationship graph until there is no path with the same starting node and end node in the initial entity relationship graph, and the optimization of the initial entity relationship graph is completed to obtain the first entity relationship graph.

[0135] The voucher data calculation module 203 is used to construct the coordinates of the historical market data according to the market entity location data and the market entity origin of the target area, and to construct the voucher data set of the target area in combination with the first entity relationship diagram and the market behavior data.

[0136] In this embodiment, the voucher data calculation module 203 includes: a voucher data calculation unit;

[0137] The voucher data calculation unit is used to obtain the location data of each market entity in the target area;

[0138] Taking the origin of the market entity as the coordinate origin, combined with the location data of each market entity, the market entity location coordinate system of the target area is constructed to obtain the coordinate information of each market entity;

[0139] Calculate the mapping value of each node, edge and path in the first entity relationship graph according to the coordinate information of each market entity;

[0140] The timestamp information and behavior type of each market behavior are obtained, and the voucher data set of the target area is constructed by combining the mapping values ​​of each node and path in the first entity relationship diagram.

[0141] In this embodiment, the credential data calculation unit includes: a mapping value calculation subunit;

[0142] The mapping value calculation subunit is used to add the horizontal coordinate value and the vertical coordinate value of each market entity and then take the absolute value to obtain the mapping value of the node corresponding to the market entity;

[0143] Get the coordinate information of the nodes at both ends of each edge, calculate the Euclidean distance between the nodes at both ends, and get the mapping value of the edge;

[0144] According to a preset path mapping function, the mapping value of each node and edge is combined to obtain the mapping value of each path in the first entity relationship graph;

[0145] The preset path mapping function includes:

[0146] ,

[0147] in, is the mapping value of the path, is the number of edges in the path, is the mapping value of each edge, is the sum of the mapping values ​​of the nodes and edges on the path, is the mapping value of the starting node of the path, is the mapping value of the end node of the path, and It refers to performing a base-10 logarithm operation on the mapping values ​​of the starting node and the end node of the path.

[0148] In this embodiment, the credential data calculation unit includes: a credential value calculation subunit;

[0149] The credential value calculation subunit is used to obtain the timestamp information of each market behavior, and perform format conversion on the timestamp to obtain the first timestamp of each market behavior;

[0150] According to a preset label assignment method, assign a value to the behavior type of each market behavior, so that each market behavior has a behavior type value;

[0151] According to the behavior type value of each market behavior, the first behavior type value of each path is obtained;

[0152] Generate an initial voucher value corresponding to each path according to a preset voucher data generation function, combining the first timestamp of each market behavior and the mapping value of each node and path in the first entity relationship graph;

[0153] The preset credential data generation function includes:

[0154] ,

[0155] in, is the credential data corresponding to the path. is the mapping value of the starting node of the path, is the mapping value of the end node of the path, is a natural constant, is the mapping value of the path, The first timestamp of the market behavior corresponding to the path; To perform a second root operation on a value;

[0156] According to a preset splicing method, the initial credential value of the path and the first behavior type value are spliced ​​to obtain the first credential value of the path, and then the credential data set of the target area is obtained.

[0157] The risk assessment module 204 is used to assess the market risk of the target area based on the voucher data set.

[0158] In this embodiment, the risk assessment module 204 includes: a risk assessment unit;

[0159] The risk assessment unit is used to classify the credential data set according to the credential data set in combination with a preset classification method;

[0160] According to the classification results, the voucher data set is divided into blocks and counted to obtain a number of risk scores corresponding to each first voucher value in the voucher data set;

[0161] Obtain preset risk assessment criteria, and evaluate the market risk of the target area in combination with the risk score.

[0162] This embodiment obtains historical market data of the target area, and obtains the market entity data set and the market entity origin through the classification model, making full use of the classification model's efficient processing capabilities for massive data, thereby efficiently obtaining a large number of different types of market entities; then, by constructing a first entity relationship diagram between the market entity data set and the market behavior data, a preliminary connection between the market entity and the market behavior is established, and combined with the market entity location of the target area, the coordinates of the historical market data are constructed, thereby constructing a voucher data set for the target area; the voucher data set constructed with actual location data enhances the correlation with the actual entity, thereby making the digital twin of the historical market data more accurate, ensuring that the actual market risk assessment of the target area based on the voucher data set is more accurate, and ensuring the market stability of the region.

[0163] Embodiment 3

[0164] The embodiments of the present invention provide a terminal device and a computer-readable storage medium.

[0165] The terminal device includes: one or more processors; a memory coupled to the processor and used to store one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the digital twin method of market data as described in the above embodiments.

[0166] The computer-readable storage medium stores a computer program thereon, wherein the computer program is executed by a processor to implement the digital twin method for market data as described in the above embodiment.

[0167] To summarize, this embodiment obtains the historical market data of the target area, and obtains the market entity data set and the market entity origin through the classification model, making full use of the classification model's efficient processing capabilities for massive data, thereby efficiently obtaining a large number of different types of market entities; then, by constructing a first entity relationship diagram between the market entity data set and the market behavior data, a preliminary connection between the market entity and the market behavior is established, and combined with the market entity location of the target area, the coordinates of the historical market data are constructed, thereby constructing a voucher data set for the target area; the voucher data set constructed with actual location data enhances the correlation with the actual entity, thereby making the digital twin of the historical market data more accurate, ensuring that the actual market risk assessment of the target area based on the voucher data set is more accurate, and ensuring the market stability of the region.

[0168] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A digital twin method for market data, characterized in that: include: Obtain historical market data of the target area and classify it in combination with the preset classification model to obtain the market entity data set and market entity origin; According to a preset search algorithm, combined with the market subject data set, the market behavior data corresponding to each market subject is obtained from the historical market data, and a first entity relationship diagram between the market subject data set and the market behavior data is constructed; Obtain location data for each market player in the target area; Taking the origin of the market entity as the coordinate origin, combined with the location data of each market entity, the market entity location coordinate system of the target area is constructed to obtain the coordinate information of each market entity; Calculate the mapping value of each node, edge and path in the first entity relationship graph according to the coordinate information of each market entity; Obtaining the timestamp information and behavior type of each market behavior, and combining the mapping values ​​of each node and path in the first entity relationship diagram to construct a voucher data set for the target area; The step of calculating the mapping value of each node, edge, and path in the first entity relationship diagram according to the coordinate information of each market entity specifically includes: Add the horizontal coordinate value and the vertical coordinate value of each market entity and take the absolute value to obtain the mapping value of the node corresponding to the market entity; Get the coordinate information of the nodes at both ends of each edge, calculate the Euclidean distance between the nodes at both ends, and get the mapping value of the edge; According to a preset path mapping function, the mapping value of each node and edge is combined to obtain the mapping value of each path in the first entity relationship graph; The preset path mapping function includes: , in, is the mapping value of the path, is the number of edges in the path, is the mapping value of each edge, is the sum of the mapping values ​​of the nodes and edges on the path, is the mapping value of the starting node of the path, is the mapping value of the end node of the path, and It means to perform a logarithm operation with base 10 on the mapping values ​​of the starting node and the end node of the path; The acquiring of the timestamp information and behavior type of each market behavior and combining the mapping values ​​of each node and path in the first entity relationship diagram to construct a voucher data set for the target area specifically includes: Obtaining timestamp information of each market behavior, converting the format of the timestamp to obtain the first timestamp of each market behavior; According to a preset label assignment method, assign a value to the behavior type of each market behavior, so that each market behavior has a behavior type value; According to the behavior type value of each market behavior, the first behavior type value of each path is obtained; Generate an initial voucher value corresponding to each path according to a preset voucher data generation function, combining the first timestamp of each market behavior and the mapping value of each node and path in the first entity relationship graph; The preset credential data generation function includes: , in, is the credential data corresponding to this path, is the mapping value of the starting node of the path, is the mapping value of the end node of the path, is a natural constant, is the mapping value of the path, The first timestamp of the market behavior corresponding to the path; To perform a quadratic root operation on a value; According to a preset splicing method, the initial credential value of the path and the first behavior type value are spliced ​​to obtain the first credential value of the path, and then the credential data set of the target area is obtained; According to the voucher data set, the market risk of the target area is evaluated; wherein, the market risk of the target area is evaluated according to the voucher data set, specifically comprising: classifying the voucher data set according to the voucher data set in combination with a preset classification method; dividing the voucher data set into blocks and performing statistics according to the classification results to obtain a number of risk scores corresponding to each first voucher value in the voucher data set; obtaining a preset risk assessment standard, and evaluating the market risk of the target area in combination with the risk score.

2. A digital twin method for market data according to claim 1, characterized in that: The historical market data of the target area is obtained, and classified in combination with a preset classification model to obtain a market entity data set and a market entity origin, specifically including: Obtain historical market data of the target area, clean the historical market data and extract market entity types to obtain initial historical market data and market entity types; Inputting the initial historical market data into a preset classification model, and combining the market subject types to obtain a plurality of classification sets output by the classification model; According to the market entity type, market entities are extracted from each classification set to obtain a market entity data set; Calculate the feature points of each classification set separately to obtain several feature points; According to the preset origin calculation method, the market entity origin is calculated in combination with the several characteristic points.

3. A digital twin method for market data according to claim 1, characterized in that: The method of obtaining the market behavior data corresponding to each market entity from the historical market data according to the preset search algorithm and combining the market entity data set, and constructing a first entity relationship diagram between the market entity data set and the market behavior data, specifically includes: According to the preset search algorithm, combined with each market entity in the market entity data set, the behavior data corresponding to each market entity is obtained from the historical market data, and a behavior information table corresponding to each market entity is constructed; According to the knowledge graph algorithm, a node is constructed for each market entity in turn, and an initial entity relationship diagram of the market entity data set and the market behavior data is constructed in combination with the behavior information table of each market entity; According to the depth-first algorithm, the initial entity relationship graph is traversed to find several paths; Each path is detected in turn. When the starting node and the end node of the path are the same, the path is removed from the initial entity relationship graph until there is no path with the same starting node and end node in the initial entity relationship graph, and the optimization of the initial entity relationship graph is completed to obtain the first entity relationship graph.

4. A digital twin device for market data, characterized in that: include: Market entity acquisition module, entity relationship diagram construction module, voucher data calculation module and risk assessment module; The market subject acquisition module is used to acquire the historical market data of the target area, and classify it in combination with a preset classification model to obtain a market subject data set and a market subject origin; The entity relationship diagram construction module is used to obtain the market behavior data corresponding to each market entity from the historical market data according to the preset search algorithm and in combination with the market entity data set, and to construct a first entity relationship diagram of the market entity data set and the market behavior data; The voucher data calculation module is used to obtain the location data of each market entity in the target area; using the location of the market entity origin as the coordinate origin, combined with the location data of each market entity, construct the market entity location coordinate system of the target area, and obtain the coordinate information of each market entity; according to the coordinate information of each market entity, calculate the mapping value of each node, edge and path in the first entity relationship diagram; obtain the timestamp information and behavior type of each market behavior, and construct the voucher data set of the target area in combination with the mapping value of each node and path in the first entity relationship diagram; The step of calculating the mapping value of each node, edge, and path in the first entity relationship diagram according to the coordinate information of each market entity specifically includes: Add the horizontal coordinate value and the vertical coordinate value of each market entity and take the absolute value to obtain the mapping value of the node corresponding to the market entity; Get the coordinate information of the nodes at both ends of each edge, calculate the Euclidean distance between the nodes at both ends, and get the mapping value of the edge; According to a preset path mapping function, the mapping value of each node and edge is combined to obtain the mapping value of each path in the first entity relationship graph; The preset path mapping function includes: , in, is the mapping value of the path, is the number of edges in the path, is the mapping value of each edge, is the sum of the mapping values ​​of the nodes and edges on the path, is the mapping value of the starting node of the path, is the mapping value of the end node of the path, and It means to perform a logarithm operation with base 10 on the mapping values ​​of the starting node and the end node of the path; The acquiring of the timestamp information and behavior type of each market behavior and combining the mapping values ​​of each node and path in the first entity relationship diagram to construct a voucher data set for the target area specifically includes: Obtaining timestamp information of each market behavior, converting the format of the timestamp to obtain the first timestamp of each market behavior; According to a preset label assignment method, assign a value to the behavior type of each market behavior, so that each market behavior has a behavior type value; According to the behavior type value of each market behavior, the first behavior type value of each path is obtained; Generate an initial voucher value corresponding to each path according to a preset voucher data generation function, combining the first timestamp of each market behavior and the mapping value of each node and path in the first entity relationship graph; The preset credential data generation function includes: , in, is the credential data corresponding to this path, is the mapping value of the starting node of the path, is the mapping value of the end node of the path, is a natural constant, is the mapping value of the path, The first timestamp of the market behavior corresponding to the path; To perform a quadratic root operation on a value; According to a preset splicing method, the initial credential value of the path and the first behavior type value are spliced ​​to obtain the first credential value of the path, and then the credential data set of the target area is obtained; The risk assessment module is used to assess the market risk of the target area based on the voucher data set; wherein, the assessing the market risk of the target area based on the voucher data set specifically includes: classifying the voucher data set based on the voucher data set in combination with a preset classification method; dividing the voucher data set into blocks and performing statistics based on the classification results to obtain a number of risk scores corresponding to each first voucher value in the voucher data set; obtaining a preset risk assessment standard, and assessing the market risk of the target area in combination with the risk score.

5. A terminal device, characterized in that: include: one or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the digital twin method for market data as described in any one of claims 1 to 3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer program is executed by a processor to implement the digital twin method for market data as described in any one of claims 1 to 3.